![](/img/trans.png)
[英]Deleting all files and directories except a few specified in a “list” in python
[英]Sum of all columns except the ones specified in a list in Python
每個ID我都有以下數據:
id ---- Base AE Val LT RO+ Prem AM TN T3 AR
05 0 34.34 9.42 70.68 0 0 0 0 0 0 0
108 0 43.77 0 28 0 0 0 0 0 0 0
205 0 77.64 0 32.2 0 0 0 0 0 0 0
320 0 66.24 0 59.628 0 0 0 0 0 0 0
313 0 21.66 0 21.442 0 0 0 0 0 0 0
324 0 72.37 0 701.12 0 0 0 0 0 0 0
505 0 76.057 0 43.87 0 0 0 0 0 0 0
現在,我想找到所有列的總和,除了我指定的幾列和其他列分開的列,如下所示:
id Base Val Others Total
05 34.34 70.68 9.42 114.441387
108 43.77 28 0 71.77
205 77.64 32.2 0 109.84
320 66.24 59.628 0 125.868
313 21.66 21.442 0 43.102
324 72.37 701.12 0 773.49
505 76.057 43.87 0 119.927
因此,如果我要保留的列列表:
cols_to_keep = ['Base','Val']
不屬於此列表的其他通道將在“其他”列中匯總,並且每行中的所有值總計為總計。 id是記錄的索引。
我能夠將我聲明的列保留在列表中,但是如何總結除“其他”列中的列表以外的其他列。 有人可以幫我嗎? 數據在pandas df中。
刪除不希望求和的列:
df['Others'] = df.drop(cols_to_keep, axis=1).sum(axis=1)
df['Total'] = df.sum(axis=1)
使用assign
,對於過濾器列,使用Index.difference
:
cols_to_keep = ['Base','Val']
c = df.columns.difference(cols_to_keep)
df = df[cols_to_keep].assign(Others=df[c].sum(axis=1), Total=df.sum(1))
print (df)
Base Val Others Total
id
5 34.340 70.680 9.42 114.440
108 43.770 28.000 0.00 71.770
205 77.640 32.200 0.00 109.840
320 66.240 59.628 0.00 125.868
313 21.660 21.442 0.00 43.102
324 72.370 701.120 0.00 773.490
505 76.057 43.870 0.00 119.927
In [47]: !cat b.txt | tr -s ' ' > data.txt
...: df = pd.read_csv("data.txt",sep=" ", dtype={'id':str})
...: df['Others'] = df['AE']
...: df['Total'] = df['Base'] + df['Others'] + df['Val']
...:
...: cols_to_keep=['id', 'Base', 'Val','Others','Total']
...: c = df.columns.difference(cols_to_keep)
...: df.drop(c, axis=1)
...: newDf = df.drop(c, axis=1)
...:
In [48]: newDf
Out[48]:
id Base Val Others Total
0 05 34.340 70.680 9.42 114.440
1 108 43.770 28.000 0.00 71.770
2 205 77.640 32.200 0.00 109.840
3 320 66.240 59.628 0.00 125.868
4 313 21.660 21.442 0.00 43.102
5 324 72.370 701.120 0.00 773.490
6 505 76.057 43.870 0.00 119.927
聲明:本站的技術帖子網頁,遵循CC BY-SA 4.0協議,如果您需要轉載,請注明本站網址或者原文地址。任何問題請咨詢:yoyou2525@163.com.